See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/315867278

Development of Blast Response Limits for Load-Bearing Prestressed Concrete

Panels Using Full-Scale Shock Tube Test Data

Conference Paper · April 2017

DOI: 10.1061/9780784480397.017

CITATIONS READS

0 200

3 authors, including:

Thomas Mander
University of Texas at San Antonio
18 PUBLICATIONS   147 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Thomas Mander on 25 May 2020.

The user has requested enhancement of the downloaded file.

 DEVELOPMENT OF BLAST RESPONSE LIMITS FOR
LOAD-BEARING PRESTRESSED CONCRETE PANELS
USING FULL-SCALE SHOCK TUBE TEST DATA

Thomas J. Mander, P.E.1, Michael J. Lowak, 1 and Michael A. Polcyn1

1
Baker Engineering and Risk Consultants, Inc., 3330 Oakwell Court, Suite 100;
San Antonio, TX 78218-3024; PH (210) 824-5960

ABSTRACT

Blast response limits have traditionally been used for predicting the level of
protection that structural components provide to buildings susceptible to explosion
effects. These limits are usually derived from a combination of post-incident damage
indicators, test data, and supporting dynamic analysis. Such limits do not exist for
load-bearing prestressed wall panels, largely due to the lack of available full-scale test
data.

This paper presents design response limits for load-bearing prestressed

concrete wall panels based on a series of full-scale shock tube tests and dynamic
analysis. The panels tested included 6-inch thick solid prestressed concrete panels
and prestressed concrete insulated panels with 3-inch thick wythes separated with
two inches of rigid insulation. Static axial loads were sustained on the top of the
panels in the dynamic shock tube tests, ranging from 5 and 10 percent of the gross
static axial capacity of each wall. The proposed limits are intended for use with non-
linear single degree of freedom analysis methods.

INTRODUCTION

Precast concrete load-bearing (LB) walls have infrequent use in blast settings,
owed to the lack of research in this area. Using LB walls in new construction, or
assessing the damage of an existing building for given blast scenarios, requires
established response criteria for a structural assessment. Presently, no criteria exists
for LB prestressed concrete walls in blast guidelines. Single-degree-of-freedom
(SDOF) response criteria, which will allow the expanded use of LB structural panels
in blast-resistant construction, are proposed by the authors.

Proposed criteria were developed through analysis and validated with a series
of shock tube tests performed on full-scale LB panels. Panel construction included
solid prestressed panels and prestressed insulated panels. LB panels supported a

1
concentric axial load with a magnitude expressed as an equivalent uniform stress that
is a fraction of the specified concrete strength of 5 percent and 10 percent of each
panel (0.05 f’cAg and 0.10f’cAg). This load conservatively represents a combination of
static (dead and live) loads and dynamic reactions from supporting components (such
as roof members). This paper presents the experimental results, which were
imperative in the development of the proposed response criteria. Analytical methods
for establishing the response criteria are published elsewhere (Mander, 2017).

RESPONSE CRITERIA FOR BLAST DESIGN

The development of blast response criteria is significantly different from the

development of standards for conventional design loads. Blast performance criteria
assume a single loading event, allowing structural damage to ensue. The extent of
damage is related to the level of protection the structure is required to provide to the
building occupants. Unlike seismic performance criteria, blast performance criteria
are based on the performance of the structural component through only one or two
cycles. Therefore, a loss of capacity may be tolerable so long as the damaged
element can still carry sustained loads after the event.

Blast analysis and design of blast loaded wall panels are most commonly
performed at a component level using SDOF methods. The peak deflection and
corresponding support rotation from the SDOF analysis are of key importance, and
are compared to prescribed limits to satisfy a level of component damage. Limiting
values are in terms of support rotation, θ, and displacement ductility, μ. Assuming a
plastic hinge for a simply-supported panel at mid-span, these values can be calculated
from the SDOF analysis results using Equation (1) and (2), respectively.

2∆
= Equation (1)
∆
= ∆ Equation (2)

where ∆max is the peak SDOF displacement, ∆y is the yield displacement, and L is the
flexural span length of the component.

ASCE 59-11 (ASCE 2011) adopts response limits from the U.S. Army Corps
of Engineers (USACE, 2008) Protective Design Center response criteria for SDOF
components. These limits were developed for Department of Defense facilities
designed against high explosive terrorist threats. Four Levels of Protection (LOP) are
defined as: High (HLOP), Medium (MLOP), Low (LLOP) and Very Low (VLLOP).
These LOPs respectively correspond to expected element damage denoted by the
USACE as Superficial, Moderate, Heavy, and Hazardous.

Currently, quantitative response criteria for LB prestressed concrete walls are

not published, motivating the need for the research presented herein. Existing

2
response limits for non-load-bearing (NLB) panels are listed in Table 1. The
prestressing index, ωp, is calculated using Equation (3).

= Equation (3)

where Aps is the prestressing area, dps is the depth to centroid of prestressing, fps is the
prestressing stress at ultimate, f′c is the concrete compression strength and b is the
width of the compression block.

Table 1. Non-Load-Bearing Prestressed Concrete Response Limits

(from PDC TR-06-08)
Superficial Moderate Heavy Hazardous
Reinforcement Index Damage Damage Damage Failure
µ θ µ θ µ θ µ θ

ωp > 0.30 0.7 - 0.8 - 0.9 - 1

0.15 ≤ ωp ≤ 0.30 or
members with 0.25 0.29 0.33
0.8 - 1° 1.5° 2°
ωp ≤ 0.15 and no shear
reinforcement
ωp ≤ 0.15 with shear
1 - - 1° - 2° - 3°
reinforcement

SHOCK TUBE TESTING

Specimen Descriptions and Setup

A total of eighteen different precast, prestressed panel specimens were

dynamically loaded with simulated blast loads in the BakerRisk shock tube. As
shown in Figure 1, all specimens were full-scale 4-foot wide panels spanning 16 feet
between supports, representative of typical interstory building heights. The otherwise
open portions of the shock tube on either side of the panel specimen’s vertical edges
were covered with stiffened steel plate bolted to the shock tube frame. This rigid
plate contains the blast wave within the shock tube, eliminating the possibility of blast
wave clearing, which would reduce the applied impulse on the wall. (Geng, 2014). A
slight (¼-inch) gap was left between the vertical edges of the panels to prevent any
form of contact with the test frame under dynamic response.

Figure 1(b) shows an LB specimen with the axial load apparatus at the top of
the wall. The axial load applicator is the first known of its kind. Traditionally,
researchers use hydraulic actuators, but these do not respond fast enough under
dynamic loading to maintain a constant axial load. The apparatus shown utilizes air
bladders that are contained within a fixed steel chamber, which applies force to
vertical steel pistons. The pistons apply load to a spreader beam, in turn loading the

3
top of the wall. This method was effective at maintaining an axial load within 10% of
the original applied axial load, even with support rotations reaching up to 4 degrees.

The axial load was concentrically applied in all tests. While some walls will
support eccentric axial loads on the interior of a building, this eccentricity counteracts
exterior blast loading. Therefore, the axial load setup utilized in the test program is
considered to be conservative for realistic loading conditions. Lateral connections
were simple bearing connections to eliminate connection variability from the dynamic
response. It was impractical to consider different connections in this study.
Additionally, a separate shock tube test program on conventionally reinforced precast
non-load-bearing panels, with various precast connections, saw panel support
rotations reach 12 degrees without observing connection failure (Bazan, 2015).
These connections were designed using LRFD methods, using design forces
associated with the ultimate resistance of the panels.

(a) Non-Load-Bearing Wall (b) Load-Bearing Wall

Figure 1. Wall Specimens Mounted in Shock Tube

Table 2 describes the panel specimens used for the shock tube test matrix.
Some panels were tested multiple times to determine different damage thresholds.
Panels were cast by certified PCI precasters. Insulated panels were constructed with a
bond breaker between the concrete wythes and insulation. This was purposely done
to conservatively simulate potential long term bond loss or poor cohesion (oil on
foam) in casting. Concrete with a compressive strength of 5000 psi was specified for
all panel specimens. Compression strengths were measured at the time of shock tube
testing. The average compression strength of three 6-inch × 12-inch concrete
cylinders is reported in Table 2. Note that the axial load applied in the shock tube is
based on the specified 5000 psi compressive strength.

4
 Table 2. Panel Specimens for Shock Tube Tests
Axial f’c at test
Panel Type Panel Construction
Load date (psi)
None 7200 6-inch thick panel with five 3/8-inch dia.
Solid Prestressed 0.05f’cAg 5200 Gr 270 strands at mid-depth (ωp = 0.15) and
0.10f’cAg 7200 WWR 6 × 6-D4 × D4 at mid-depth.
3/2/3 panel, with three 3/8-inch dia. Gr 270
None 7000 strands (ωp = 0.04) and WWR 6 × 6-D4 × D4
Prestressed Insulated mid-depth in each wythe.
Fully Composite 0.05f’cAg 7100 Two continuous P12G welded wire girders with
(for Ultimate Strength) 0 gauge (0.306 inch) top and bottom wire,
0.10f’cAg 7000 3 gauge diagonal (0.243 inch),
5.5 inch truss height.
None 6900
Prestressed Insulated Truss girders spaced intermittently to equate to
Partially Composite 0.05f’cAg 5200 60% of shear connectors provided for the fully
(for Ultimate Strength) composite design.
0.10f’cAg 6900

Experimental Results of Shock Tube Tests

The experimental data for solid prestressed panels, fully composite panels,
and partially (60%) composite panels are provided in Table 3, Table 4, and Table 5,
respectively. The results are divided into panel type, rather than in test number order.
Although only eighteen panel specimens were available, thirty shock tube tests were
completed. Panels that were tested multiple times are indicated with an asterisk. The
tables include peak pressure, applied impulse, peak mid-height displacement, ∆max,
peak support rotation, θmax, residual mid-height displacement, ∆res, residual support
rotation, θres, and a qualitative description of damage.

Qualitative damage was defined in terms of visible damage. Superficial damage

was specified for panels with hairline cracking and no permanent displacement.
Moderate damage included panels with cracks less than ¼-inch wide (typical
acceptable width for epoxy injection repair) and maximum residual displacements
less than L/360 for LB panels, where L is the panel span length. This residual
displacement equates to a ½-inch out-of-plane mid-height displacement for the panels
tested. It was selected based on the allowable PCI Design Handbook (Chapter 13.2.8)
bowing tolerances, recognizing that Moderate damage implies a component could be
re-used. Finally, Heavy damage was classified as visible damage exceeding the
Moderate threshold.

5
 Table 3. Solid Prestressed Panel Shock Tube Test Results
Axial P i ∆max ∆res
Test θmax θres Observed Damage
(kips) (psig) (psi-ms) (inch) (inch)
1 0 4.4 78 1.8 1.1° 0.2 0.12° Hairline Cracking (Superficial)

2* 0 5.9 103 4.2 2.6° 0.4 0.25 Cracking & residual disp. (Moderate)

3* 0 5.9 112 5.9 3.6° 0.5 0.31° Widespread Cracking (Heavy)

4* 0 7.2 146 Fails ~12 7.4° - - Strand Fracture (Blowout)

10 144 6.0 114 2.3 1.4° 0.1 0.06° Cracking & residual disp. (Moderate)

11* 144 6.6 127 3.1 1.9° 0.3 0.18° Cracking & residual disp. (Moderate)

15 144 7.0 128 3.4 2.1° 0.44 0.27° Cracking & residual disp. (Moderate)

16* 144 7.3 135 Fails ~5.5 3.4° - - Mid-height comp. failure (Blowout)

21 88 6.4 105 2.6 1.6° 0.19 0.12° Cracking & residual disp. (Moderate)

22 72 7.0 120 3.6 2.2° 0.25 0.15° Cracking & residual disp. (Moderate)

27 72 5.9 105 2.75 1.7° 0.1 0.06° Cracking & residual disp. (Moderate)

28* 72 7.1 122 3.9 2.4° 0.63 0.39° Significant residual disp. (Heavy)
* Denotes previously tested specimen

Table 4. Fully Composite Prestressed Insulated Panel Shock Tube Test Results
Axial P i ∆max ∆res
Test θmax θres Observed Damage
(kips) (psig) (psi-ms) (inch) (inch)
5 0 6.4 122 2.3 1.4° 0.8 0.5° Cracking & residual disp. (Moderate)

6* 0 6.9 138 5.1 3.1° 1.5 0.9° Prestress bond failure cracks (Heavy)

12 144 7.1 148 Fails ~ 7 4.3° - - Mid-height strand fracture (Blowout)

17 144 4.5 75 1.3 0.8° 0 0° No visible cracks (Superficial)

18* 144 5.6 93 1.9 1.2° 0.44 0.3° Cracking & residual disp. (Moderate)
19* 144 6.4 115 3.3 2.0° 1.1 0.7° Significant permanent disp. (Heavy)

23 72 7.2 130 2.0 1.2° 0.31 0.2° Cracking & residual disp. (Moderate)

24 72 7.8 147 2.9 1.8° 0.75 0.5° Significant residual disp. (Heavy)

29 72 8.7 168 3.9 2.4° 1.25 0.8° Significant residual disp. (Heavy)
* Denotes previously tested specimen

6
 Table 5. Partially Composite Prestressed Insulated Panel Shock Tube Results
Axial P i ∆max ∆res
Test θmax θres Observed Damage
(kips) (psig) (psi-ms) (inch) (inch)
7 0 4.2 79 1.5 0.9° 0.3 0.2° Cracking & residual disp. (Moderate)

8* 0 6.4 125 4.3 2.6° 0.9 0.6° Significant residual disp. (Heavy)

9* 0 7.1 150 7.0 4.3° 1.3 0.8° Significant residual disp. (Heavy)

13 144 4.4 77 1.5 0.9° 0 0° Hairline Cracking (Superficial)

14* 144 5.3 97 2.5 1.5° 0.5 0.3° Cracking & residual disp. (Moderate)

20 144 6.8 123 3.3 2.0° 1.1 0.7° Significant residual disp. (Heavy)

25 72 7.0 120 2.4 1.5° 0.4 0.2° Cracking & residual disp. (Moderate)

26 72 7.9 144 3.2 2.0° 0.9 0.5° Significant residual disp. (Heavy)

30 72 8.1 157 5.5 3.4° 2.0 1.2° Cracking & large perm. def. (Heavy)
* Denotes previously tested specimen

PROPOSED LOAD-BEARING RESPONSE LIMITS

Blast response criteria are proposed for solid prestressed, and double-wythe
prestressed insulated wall panels. Response limits were derived using analytical
moment-curvature models, which were validated with shock tube testing. These
models accounted for material non-linearity to generate resistance functions
exhibiting hardening and softening regions. Unlike traditional elastic-perfectly-
plastic (EPP) resistance functions commonly used in blast design, this method
provides insight to actual material strains and corresponding damage. The analytical
models were used in a parametric study to determine response criteria limits suitable
for typical prestressed wall construction. Full details of these models can be found
elsewhere (Mander, 2016 and Mander, 2017).

As blast analysis is customarily performed using EPP resistance functions, it

is necessary to propose response criteria applicable for such methods. Figure 2
illustrates the process of converting the moment-curvature derived response criteria
into criteria that is applicable to EPP resistance functions. Resistance functions
(Figure 2a) are calculated using both methods, and the support rotation limits are
calculated for the moment-curvature model (Figure 2b). For each damage threshold,
the strain energy, U, is calculated (Figure 2c), and an equivalent support rotation is
calculated using the elastic-plastic model by preserving strain energy (Figure 2d).
The moment-curvature thresholds were set by levels of strain in the concrete and
prestressing strands, as detailed in Table 6.

7
(a) Compare M-Ø and EPP Resistance (b) Determine M-Ø Damage Thresholds
Functions

∅ =

∅
=

(c) Calculate Equivalent Strain Energy (d) Convert Support Rotation to EPP
Figure 2. Process for Developing Response Criteria for
EPP Resistance Functions

Table 6. Selected Damage Limit States for Analytical Model

USACE
PDC TR-06-08 BakerRisk
Component
Component Consequence Selected Threshold
Damage
Prestressing below proportional limit of
0.0085. Strains below this level will cause
Superficial Component has no visible damage
near elastic response, with cracks closing after
event. Permanent displacement not visible.
Component has some permanent The smaller of prestressing at yield threshold
deflection. It is generally of 0.011, which will cause permanent
Moderate repairable, if necessary, although displacement, or concrete reaching its peak
replacement may be more stress, f’c. Load-bearing panels with
economical and aesthetic permanent displacement less than L/360.
Component has not failed, but it
Concrete reaching a strain of 0.003, associated
has significant permanent
Heavy with concrete crushing. Load-bearing panels
deflections, causing it to be
resistance dropping to 0.8Ru.
irreparable.
Component has failed, and debris
Resistance dropping to 0.8Ru, classified as
Hazardous velocities range from insignificant
failure point
to very significant.

8
 To capture geometric P-delta effects, a ductility limit threshold was set for LB
wall components. This limit, given in Equation 4, was derived by including P-delta
effects in the moment-curvature resistance function as shown in Figure 3 and
computing the ductility at which the resistance drops to 0.8Ru. Strain energy was
preserved from the moment-curvature analysis from which the ductility limit was
calculated using an EPP resistance function. Once this value was computed for
multiple prestressing indexes, Equation 4 was best fit to the data. The span-to-depth
ratio was incorporated into this equation to recognize that slender sections are more
prone to P-delta failures.

.
=
∆ /
Equation (4)

The ductility equation becomes applicable for walls with axial loads that are
significant compared to the moment capacity of a wall in flexure alone. For example,
a wall is more susceptible to be governed by P-delta effects when it has a low
prestressing index and a long span (low flexural resistance).

Resistance Function without P-delta Effects Resistance Function Including P-delta Effects
Figure 3. Determining Failure Limit from P-delta Moment

LOAD-BEARING SOLID PRESTRESSED PANELS

Axial load levels of 0.05f’cAg and 0.10f’cAg were considered in the analysis of
LB solid panels. Limits were set using the same damage states of Table 6 from the
analysis performed and shock tube test data. Heavy and Hazardous support rotations
were set equal, as a LB failure is a brittle failure, with a sudden transition from
concrete crushing to complete loss in load-carrying capacity. Proposed response
limits are listed in Table 7. The proposed limits are plotted as horizontal lines in
Figure 4 versus the observed damage for all load bearing tests.

9
 Table 7. Proposed Response Limits for Solid Prestressed LB Wall Panels
Superficial Moderate Heavy Hazardous
Wall Type
µ θ µ θ µ θ µ θ
Load-bearing 1 1° 1.5° 2° 2°

Figure 4. Observed Damage and Proposed Limits for Solid Prestressed Panels

FULLY AND PARTIALLY COMPOSITE PRESTRESSED INSULATED

PANELS

Axial load levels of 0.05f’cAg and 0.10f’cAg were considered for LB insulated
double wythe wall panels. This axial load was assumed to act concentrically with
equal magnitude on both wythes. Full composite action (stiffness and strength) was
assumed for the parametric analysis performed (Mander, 2017). Owing to the fact
that the response values are based on peak strains, the proposed limits will be
conservative for partially composite panels, such as the partially composite panels
tested dynamically in this research. Proposed response limits are listed in Table 8.
The proposed limits are plotted as horizontal lines in Figure 5 versus the observed
damage for all fully and partially composite load-bearing tests.

Table 8. Proposed Response Limits for Prestressed Insulated LB Wall Panels

Superficial Moderate Heavy Hazardous
Wall Type
µ θ µ θ µ θ µ θ
Load-bearing 1 0.5° 1.5° 2° 2°

10
 Fully Composite Partially Composite
Figure 5. Observed Damage and Proposed Limits for Solid Prestressed Panels

Discussion on Proposed Criteria

The proposed LB limits in Table 7 and Table 8 can be compared with the
existing blast NLB limits in Table 1. It is evident that the proposed LB limits actually
exceed the existing response limits for typical wall panels (ωp ≤ 0.15 and no shear
reinforcement). In order for the proposed results to therefore be implemented
subjectively, NLB limits must be revised. Based on the limited shock tube tests on
NLB panels performed by the authors, limits have been proposed elsewhere (Mander,
2017).

The proposed criteria are applicable for simply supported walls that are
controlled by flexural failure (adequate shear capacity) and far-range blast effects.
The limits should be used in conjunction with SDOF resistance functions computed
using the PDC TR-06-01 methodology (USACE, 2006). For LB panels, the effect of
axial load on the compression block depth and increased moment capacity must be
accounted for in the resistance function. P-delta effects can be incorporated in the
resistance function, or as an equivalent lateral load. The response limits are
applicable for static and dynamic axial compression loads. Continuous panels must
consider shear-flexure interaction, which was outside of the scope of this research.

CONCLUSIONS AND FUTURE RECOMMENDATIONS

This research demonstrated the viability of using load-bearing wall panels for
blast-resistant design. A total of thirty shock tube tests were completed on various
16-foot tall prestressed wall panels. The walls tested included solid prestressed walls,
and fully- and partially-composite walls. LB panels supported static axial loads of
0.05f’cAg and 0.10f’cAg. These specimens were able to achieve 2º of support rotation
without failure. Failure was typically governed by geometric P-∆ effects, rather than
concrete crushing or strand fracture. This stressed the importance for P-∆ effects to
be considered in the dynamic response of thin LB elements with low resistances.

11
 Proposed limits set a threshold for Moderate damage at 1.5º and Heavy damage at 2º.
A ductility term is also proposed to address P-delta failure mechanisms. These limits
are only applicable for simply supported walls subjected to a uniform blast load, and
not close-in effects. The walls must also be controlled by flexural failure, rather than
shear.

FUTURE RESEARCH
The insulated panels that were tested utilized steel truss systems connecting
the wythes. It is recognized that a variety of proprietary shear connectors are
available to the precast industry. It is recommended that these systems be tested to
ensure they satisfy the proposed response limits.

ACKNOWLEDGMENTS

This work was funded by the Precast/Prestressed Concrete Institute. A select

advisory committee provided valuable guidance throughout the course of this work,
specifically, Roger Becker, Greg Force, Suzanne Aultman, Phil Benshoof, Steven
Brock, James Davidson, John Geringer, John Hoemann, Clay Naito and Pat O’Brien.
Recognition is also given to Coreslab Structures (Texas) Inc. and Tindall Corporation
(Texas) for fabricating and supplying the prestressed wall specimens.

REFERENCES

ASCE 59-11, (2011). Blast Protection of Buildings, Published by the American

Society of Engineers (ASCE), Reston, VA.
Bazan, M., and Oswald, C. (2015). “Testing and Analysis of Precast Concrete Wall
Panels,” 16th International Symposium for the Interaction of the Effects of
Munitions with Structures, Destin, FL.
Geng, J., Mander, T.J., and Baker, Q.A., (2014). “Blast Wave Clearing Behavior for
Positive and Negative Phases,” Journal of Loss Prevention in the Process
Industries.
Mander, T.J., Lowak, M.J., and Polcyn, M.A. (2016). “Blast Performance of Load-
Bearing and Non-Load-Bearing Prestressed Concrete Panels,” 2016 PCI
Convention and National Bridge Conference, Nashville, TN.
Mander, T.J., Lowak, M.J., and Polcyn, M.A. (2017). “Blast Response Limits for
Load-Bearing Prestressed Concrete Panels,” 2017 Precast/Prestressed
Concrete Institute (PCI) Convention and National Bridge Conference,
Cleveland, OH.
USACE (2006). PDC TR-06-01, “Methodology Manual for the Single-Degree-of-
Freedom Blast Effects Design Spreadsheets (SBEDS),” U.S. Army Corps of
Engineers (USACE), Protective Design Center (PDC) Technical Report 06-
01.
USACE (2008). PDC TR-06-08, “Single Degree of Freedom Response Limits for
Antiterrorism Design,” USACE PDC Technical Report 06-08.

12

View publication stats